Testcase failed due to real ceph cluster
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Cinder |
Triaged
|
Low
|
Eric Harney | ||
Victoria |
Triaged
|
Low
|
Unassigned | ||
Wallaby |
Triaged
|
Low
|
Eric Harney |
Bug Description
Description
===========
I cloned the cinder repository of victoria branch and ran
tox tests. Unfortunately, the testcase which is 'cinder.
drivers.
failed. This testcase was affected by the real ceph cluster which
had already produced the file whose path is '/etc/ceph/
admin.keyring'. However, testcases should be isolated instead of
affecting by their execution environments.
Steps to reproduce
==================
1. Deploy a ceph cluster on a node and make sure /etc/ceph/
admin.keyring exits.
2. Clone the cinder repository of victoria branch and ran
command such as 'tox -e py3'
Expected result
===============
All testcases succeed.
Actual result
=============
The testcase which is 'cinder.
drivers.
failed.
Logs & Configs
==============
=======
Failed 1 tests - output below:
=======
cinder.
-------
Captured traceback:
~~~~~~~~~~~~~~~~~~~
Traceback (most recent call last):
File "/root/
self.
File "/root/
self.
File "/root/
raise mismatch_error
testtools.
reference = None
actual = '''\
[client.admin]
\tkey = AQAEFtRfl81/
\tcaps mds = "allow *"
\tcaps mgr = "allow *"
\tcaps mon = "allow *"
\tcaps osd = "allow *"
'''
tags: | added: ceph unit-test |
Changed in cinder: | |
status: | New → Triaged |
importance: | Undecided → Low |
Changed in cinder: | |
assignee: | nobody → Eric Harney (eharney) |
Issue was introduced by commit be02e0f3467b8cac in victoria